It is believed that much of the clinical variability among patients with sickle cell disease (SCD) may be genetically determined, resultingfrom the co-inheritanceof "epistatic"genes which interact with the basic sicklingdefect to modifythe disease pathophysiology.We recentlyconducted a pilotstudy comparing the inheritanceof several commonbloodgroup polymorphisms with SCD severity in 103 children and found that the Lewisnegative Le(a-b-) phenotype was associatedwith a 2-fold higher hospitalization rate for SCD-related complicationscompared to Le(a+b-) and Le(a-b+) patients. The Le(a-b-) phenotypeis also known be associatedwith a 2-fold increased riskof ischemic heart disease (IHD). No mechanism has yet been identified,but we hypothesize that the association between Lewis RBC phenotype and SCD severitymay be mediated by differencesin the plasmalevels of sialyI-Lewisa (sLea), a high-affinity selectin ligand.The objectives of this proposedmulti-centercollaborativestudy are: a) to confirmour initial findings in a larger group of children drawn from multiple centers; b) to determine whether the .-_ Lewis(a-b-) phenotype is also associated with disease severity in adults; c) to determine whether the Lewis phenotype or plasma sLea level predict specific types of complication (e.g. stroke, ACS); d) to establish whether Lewis acts independently of HbF, beta-globin haplotypes and alpha thalassemia; e) to look for any link between other blood group polymorphisms (e.g., Duffy, MNS) and SCD disease severity. Lewis antigen status will be determined serologically, along with 20 other blood group antigens. We will also perform Se and Le genotyping by PCR-RFLP and quantify plasma sLea by ELISA. Disease severity data will be collected via standardized report forms and entered into the CSCC Common Database. It is anticipated that retrospective data will be available from the Database for many patients. The participation of multiple centers in this project is essential, since it requires the collection of high-quality clinical data on large numbers of patients. Given the large increment in hospitalizations associated with Le(a-b-) in our pilot study, we believe that this marker may be useful as an early predictor of severity in SCD, and may help with the targeting of aggressive interventions (BMT, chronic transfusion) to higher-risk SCD patients.